exoALMA XIX: Confirmation of Non-thermal Line Broadening in the DM Tau Protoplanetary Disk

Turbulence is expected to transport angular momentum and drive mass accretion in protoplanetary disks. One way to directly measure turbulent motion in disks is through molecular line broadening. DM Tau is one of only a few disks with claimed detection of nonthermal line broadening of 0.25cs-0.33cs, where cs is the sound speed. Using the radiative transfer code mcfost within a Bayesian inference framework that evaluates over five million disk models to efficiently sample the parameter space, we fit high-resolution (0.15", 28 m s-1) 12CO J = 3-2 observations of DM Tau from the exoALMA Large Program. This approach enables us to simultaneously constrain the disk structure and kinematics, revealing a significant nonthermal contribution to the line width of ~0.4cs, inconsistent with purely thermal motions. Using the CO-based disk structure as a starting point, we reproduce the CS J = 7-6 emission well, demonstrating that the CS (which is more sensitive to nonthermal motions than CO) agrees with the turbulence inferred from the CO fit. Establishing a well-constrained background disk model further allows us to identify residual structures in the moment maps that deviate from the expected emission, revealing localized perturbations that may trace forming planets. This framework provides a powerful general approach for extracting disk structure and nonthermal broadening directly from molecular line data and can be applied to other disks with high-quality observations.

💡 Research Summary

This paper presents a comprehensive measurement of non‑thermal line broadening—i.e., turbulent motions—in the protoplanetary disk around the T Tauri star DM Tau. Using high‑resolution (0.15″, 28 m s⁻¹) ALMA observations of the 12CO J=3‑2 line from the exoALMA Large Program, the authors employ the radiative‑transfer code MCFOST within a Bayesian inference framework to explore a parameter space of more than five million disk models. The model includes a tapered‑edge surface density profile, self‑consistent temperature calculation, CO and CS chemistry (including freeze‑out, photodissociation, and photodesorption), dust settling, and a free non‑thermal broadening factor f_turb that scales with the local sound speed.

An affine‑invariant MCMC sampler (emcee) generates synthetic data cubes at each iteration, which are convolved with the observed beam and compared to 17 selected velocity channels that capture the bulk of the CO emission. By allowing key parameters such as stellar mass, gas mass, CO abundance, and the turbulence factor to vary, the analysis avoids the biases that plagued earlier studies which fixed many of these quantities. The posterior distribution peaks at f_turb ≈ 0.4, meaning the turbulent velocity v_turb is roughly 40 % of the local sound speed cₛ. This non‑thermal contribution is significantly larger than the previously reported 0.25–0.33 cₛ for DM Tau and clearly exceeds pure thermal broadening.

To test the robustness of this result, the CO‑derived temperature and density structure is held fixed while modeling the CS J=7‑6 line, which is more sensitive to non‑thermal motions because CS is heavier. Only the CS abundance and the fraction of gas‑phase CS in the freeze‑out region are varied, and the same f_turb ≈ 0.4 reproduces the observed CS emission, providing an independent confirmation of the turbulence level.

Subtracting the best‑fit model from the data reveals residual structures in the moment maps at specific radii (∼80 au and ∼150 au). These localized deviations could be caused by pressure bumps, vortex‑like flows, or nascent planets perturbing the gas, suggesting that the high‑resolution data can also trace dynamical signatures beyond the global turbulence.

Translating the measured turbulent velocity into the α‑parameter of the classic α‑disk model yields α ≈ 10⁻³, consistent with moderate turbulence but higher than the α ≲ 10⁻⁴ inferred from dust settling studies in the inner disk. This discrepancy implies that turbulence strength may vary radially or that multiple mechanisms (magnetorotational instability, vertical shear instability, planet‑disk interactions) operate simultaneously.

Overall, the study demonstrates that a joint Bayesian‑MCFOST approach can simultaneously constrain disk structure and non‑thermal line broadening directly from molecular line data. The methodology is scalable to other disks with high‑quality ALMA observations, offering a pathway to statistically map turbulence across a population of protoplanetary disks and to link turbulent properties with planet formation processes.